Description

I own shares of Exelon.  The stock is at $32.  It is down 65% from its all-time high in 2008, and it is the 65^th worst-performing stock in the S&P 500 over the last 52 weeks, down 16.5% while the index is up 13.2%, excluding dividends.

I think the downside is $33 by YE2015, and the upside is $51 by YE2015.  In the mean time you’re getting paid a 3.8% annual dividend, which amounts to $3.72 per share cumulatively through 2015.  The immediate downside is BVPS of $26.  If this goes to plan the total return would be 80% and the IRR would be ~20% with low risk.

You can download my model here: 

https://www.dropbox.com/s/keeol0q2j20plok/Exelon%20Valuation.xlsx

My argument rests on the notion that Exelon’s ultra-high quality merchant generation assets will benefit from higher electricity prices in the wholesale market driven by the following dynamics:

1.  LNG exports:  natural gas generally sets the marginal price of electricity.  The U.S. is producing >67 Bcf/day of natural gas; Canada is producing 10 Bcfd for a total of >77 Bcfd.  Cheniere Energy will begin exporting 2.5 Bcfd, or 3% of combined production starting in late 2015/early 2016.  The project is permitted, financed, and has a turnkey construction contract.  There are at least 15 Bcfd of additional projects on the table in the United States and a further 4 Bcfd of projects in Canada. 

Further, the economy is undergoing a structural increase in demand for gas for the first time in years.  U.S. gas consumption averaged 61.9 bcfd for the 10 years ending 2009.  Then it rose to 65.1bcfd in 2011, the first time above 63.6 in history.  Then it rose to 66.8 bcfd in 2012.  Demand growth is real and it’s happening. 

Read utah1009’s writeup on natural gas from earlier today for a well-articulated case for why prices could rise substantially.

 2.  Plant retirements:  coal plants around the country are being retired, either because they can’t compete with falling cost natural gas generation, and/or because they can’t justify the economics of investing capital to meet heightened environmental requirements.There are seven main electricity markets in the United States. 

The Pennsylvania New-Jersey (PJM) interconnection is where most of Exelon’s merchant nuclear plants are located.  There are ~180,000 Megawatts (MW) of generating capacity in that region.  It’s the only region forecasted to lose capacity net of planned additions.

 Source: SNL Energy

I have what I believe is a better forecast of PJM retirements from another merchant operator in that market.  It was built from the bottom up, plant by plant, and it shows 10,765 MW of gross plant retirements moving forward. 

Spot prices at Henry Hub are ~$3.80 and the various locational forward curves are ~$4.40-$4.80 in 2016.    I am modeling a $1.00 uplift starting in 2016.
PJM spot prices for electricity are $45-55 per Megawatt-hour (MWh) and the forward curves are $40-55/MWh in 2016.  I am modeling a $10 uplift starting in 2016.
I chose $1/mmbtu and $10/MWh because they are nice round numbers, and because they are within the zone of reasonableness. 

Heat rates represent the amount of energy (measured in mmbtu) required to turn fuel into 1 kilowatt-hour (kWh) of electricity.  Market heat rates represent the maximum mmbtu a plant can burn to make 1 kWh and achieve breakeven, given fuel input prices and power output prices.  For example, if natural gas costs $4.00 and electricity prices are $40/MWh then a natural gas-fired plant would have to have a heat rate less than 10,000 to break even on a variable cost basis. 

Implied market heat rates for PJM in 2016 are 11,000 – 11,500.  The uplift I am modeling for power and gas prices actually results in slight heat rate contraction.

If everything goes to plan, I think Exelon earns ~$3.50 in 2016 on an unhedged basis.  Put a 14.4x NTM multiple on it (the current median for the Philadelphia Utility Index, also consistent with the group’s historical multiples) at YE2015 and you get a $51 stock.  If none of the uplift occurs, and they just earn what the forward curves imply, they earn $2.32 on an unhedged basis. Again, put a 14.4x multiple on it at YE2015 and you get a $33 stock.

In the mean time, it’s trading at 12.8x the midpoint for 2013 guidance of $2.50 and 1.2x book value of $26 per share. 

Exelon systematically hedges its generation book 1-1 ½ years in advance so on a hedged basis the full impact of the uplift I am modeling would not hit earnings immediately.  But, markets are about expectations, and I think people will see what’s happening to Exelon’s earnings power as gas and power prices rise.

***

Hannibal Lector:  “And how do we begin to covet, Clarice?  Do we seek out things to covet?  Make an effort to answer now.”

Clarice Starling:  “No.  We just…”

Hannibal Lector:  “No.  We begin by coveting what we see every day”.

In a prior life I used to look at Exelon’s businesses on a regular basis.  I didn’t seek it out; I was required to follow it because of the nature of my job responsibilities, and I began to covet their stock because of what I saw every time I looked: 

• The adoration they received from investors and rating agencies

• The premium valuation multiples:  >3x book value vs. 1.6x-1.7x for peers

• The high rate of dividend growth:  11.0% CAGR 2001-2009

• The prized nature of their nuclear fleet, an asset so rare and special no competitor could ever hope to replicate it, even with access to unlimited capital; easily earning double digit returns on capital through the cycle in an industry whose economics are terrible

• These guys are truly first class.  Each nuclear plant in the United States receives an “INPO” rating from the Institute of Nuclear Power Operations.  It’s on a scale of 1-5 with 1 being the highest.  The ratings are private but some companies choose to release them.  Of the 11 plants ExGen had before they recently merged with Constellation, 10 were rated “1” and 1 was rated “2”. 

• The acumen of their management team to place those nuclear plants inside of a merchant business model at the onset of deregulation over a decade ago, thus ensuring their merchant business would have by far the lowest marginal cost of production

• The cleverness they demonstrated by unloading their main source of potential environmental liabilities - Commonwealth Edison’s (ComEd) old coal plants - on a management team from Edison International that was stricken with a case of deal fever

• The premium price they received for those coal assets; a deal that years later is still talked about with a tone of (deserved) smugness by Exelon employees who were around to see it happen

 (Post script – the former ComEd coal assets are operating under Chapter 11 bankruptcy protection as of December 2012, the victim of a one-two-three combination of falling power prices, capex requirements to meet new environmental standards, and rising “rent” payments under a convoluted sale-leaseback structure they designed in order to achieve tax deductibility of the interest portion of the payments.)

I coveted Exelon’s stock, but I never bought it, because it wasn’t cheap enough until now.  What brought the stock down from its peak was the drop in natural gas and power prices, and the inevitable dividend cut (from $2.10 to $1.24) they finally announced on their Q4 earnings call

COMPANY DESCRIPTION

Exelon is an integrated utility headquartered in Chicago.  They have two primary sources of earnings:           

(Owning both a regulated utility and merchant generation assets is what earns a power company the “integrated” classification.)

Commonwealth Edison was the original predecessor company to Exelon, formed in Chicago in 1907.  The way I’ve heard the story told, ComEd was the first utility to propose the basic industry concept of being awarded a monopoly franchise in exchange for submitting its customer rates to regulation. 

ComEd subsequently merged with PECO and they created the merchant businesses in 2000.  ExGen began operations January 1, 2001.  At the time power prices were at levels that required them to take substantial writedowns on the carrying value of their merchant fleet.  As a result, returns on capital as reported are somewhat overstated because the amount of capital in the denominator is lower than it should be, although returns are still extremely attractive even if you adjust for that.

Exelon acquired Constellation Energy last year.  The deal closed March 12, 2012.  Through that acquisition Exelon came to own Baltimore Gas & Electric, and added three merchant nuclear plants to its 11 pre-existing ones. 

It would be fair to say Exelon used to have an acquisition fetish but with the integration of the Constellation merger the company will have its hands full for a while. 

Adjusted EPS History:  

Source: company press releases   

ExGen's Capacity:

Source: SNL Energy
Note: stated capacities in the company’s annual report are higher.  I am only showing the capacities in this table that I am actually running through my model.

Exelon’s Utilities:

Source: company annual reports
*at Dec-31-2011

Other: 

 Source: Capital IQ, company annual reports

EXELON’S UTILITIES

I am modeling ~$1 per share of normalized earnings from the three utilities combined.

The utilities are not the highlight of the story.  They simply provide a base level of underlying earnings power that, while not exciting, is somewhat material to consolidated earnings and should be fairly consistent over time. 

The minimum portion of a utility’s capital structure that is required to be funded with common equity, and the ROE they are authorized to earn on that equity, are determined through regulatory proceedings.  

Whether the utility actually earns its authorized ROE is another matter.  The two basic determinants are the skill of management and the constructiveness of the regulators.  Management skill can only accomplish so much.  The regulatory environment is the primary driver.

Exelon’s utilities are in some of the less desirable jurisdictions.  S&P classifies each state’s local regulator into five buckets: 

• Most credit supportive
• More credit supportive
• Credit supportive
• Less credit supportive
• Least credit supportive

Pennsylvania (PECO) is a “credit supportive” jurisdiction.  Illinois (Commonwealth Edison) and Maryland (Baltimore Gas & Electric) are in “least credit supportive” jurisdictions.  ComEd and BG&E historically have underearned vs. their authorized ROEs.  PECO has tended to earn very high ROEs. 

I am going to vastly oversimplify a subject that that could take up many pages, and assume the three utilities will earn a normalized blended ROE of 8.5% on their common equity of $9.8 billion, which would result in earnings of ~$1 per share.  I would be glad to discuss this issue more in the Q&A but I think I am on relatively sound footing making that call.

Authorized ROEs will come down over time in line with interest rates.  I am not adjusting for that because they won’t come down quickly and because the utilities are underearning anyway.

EXELON GENERATION

I am modeling $2.85 per share of “cash” EPS on an unhedged basis in 2016.  Again, they hedge forward on a systematic basis which will delay the full benefit of rising gas prices on the actual EPS numbers they report, but I think the market will see the rising earnings power and price those expectations into the stock.

Let’s go over the assumptions in my model.  It was built on a bottoms basis using actual plant-by-plant operating and cost data from SNL Energy on every plant in ExGen’s portfolio, excluding wind and solar. 

Gas and power price assumptions:

I start with forward curves for electricity and natural gas at the locations where Exelon’s various plants are doing business.

For the natural gas curves, I add $1.00/mmbtu to each locational curve in 2016 and beyond, resulting in an average gas price among the various curves of $5.55/mmbtu for 2016 and slightly higher prices in 2017 and 2018.

For the electricity price curves, I add $10/MWh to each locational curve in 2016 and beyond, resulting in average prices of $58/MWh for the various curves in 2016 and slightly higher prices in 2017 and 2018.

The $10/MWh price expansion I assume actually results in slight heat rate contraction. 

Dispatch assumptions:

• Nuclear plant utilization of 93.4%:  Since the nuclear plants are low-cost baseload generation that run 24/7 whenever possible, I assume they run at maximum utilization rates, which I define as their historical average from 2007-2011, regardless of where gas prices are.  The average was 93.4%.
  
  I follow the same methodology for their hydro plants, assuming they run at their average utilization rates from 2007-2011 regardless of where gas prices are since the variable cost of production is so negligible.

• Gas plant utilization of 25%: I assume Exelon’s natural gas plants only run when the combination of power and gas prices at any given point in time allow them to generate positive incremental cash flow.  In my model that results in 25% utilization rates. 

• Coal plant utilization of 0%:  I assume Exelon’s coal plants generate no energy revenues, since they aren’t efficient enough to deliver positive variable margins given the forward curves I have in place.

Retail margin assumptions:

Exelon has a retail business they picked up with the Constellation acquisition.  The basic value proposition for customers of the retail business is the company they agree to buy power from will charge them a slight premium over the life of the contract in exchange for guaranteeing the customer price certainty. 

The sell side is modeling retail margins of $2-4/MWh for this business.  I am modeling $1.50 in order to more fully put the burden of the company’s earnings power on the generation assets.

Operating cost assumptions:

Nuclear costs are the main driver of ExGen’s cost structure, and there are three main cost categories for the nuclear plants: 

• Fuel:   the company’s nuclear fuel costs per MWh have grown at a 9.1% CAGR from 2007-2011.  I assume it grows 9.1% again in 2013, and levels off at 2.5%.  Uranium price expansion has leveled off after the Fukushima disaster.  Germany is phasing out its use of nuclear plants.  Demand is simply not as robust as it used to be.

• Maintenance:  ExGen’s costs per MWh have grown at a 2.9% CAGR from 2007-2011.  I assume 3.1% growth to reflect generally heightened concerns over nuclear safety in the wake of Fukushima and a series of very minor incidents at nuclear plants in the U.S. in recent years. 

• Non-fuel O&M:  costs per MWh have grown at 5.1% CAGR from 2007-2011.  Similar to my maintenance assumption, I add 20 bps to the inflation CAGR, resulting in 5.3% growth.

• Non-nuclear costs:  I assume a generic 3% fixed cost CAGR at the gas plants and the hydro plants.  In addition, I burden the P&L with the fixed costs of keeping the coal plants online even though they aren’t efficient enough to generate positive variable cash flows given the forward curves I have in place. 

Depreciation adjustments:

Depreciation expense was $768 million last year.  That run rate is artificially low because the PP&E of the plants was written down over a decade ago when they were placed in the merchant business.  Maintenance capex is actually >$1 billion.

I am adding $300 million per year of maintenance capex and expensing it immediately.  I am doing that to fully eliminate the benefit to GAAP earnings of depreciation charges being less than maintenance capex. 

Capacity revenues:

Capacity revenues are locked in through 2015 due to auctions that have already taken place.  Starting in 2016 I assume capacity revenues are held constant at 2015 levels. 

MODEL RESULTS

The decline in earnings post-2016 is just a function of assuming zero growth in the company’s top line while fixed costs continue to rise.  It’s not because I’m making a call that the earnings power will tip over and start deteriorating.  

Room for error:

• I am assigning zero value to the wind and solar assets
• I am assuming zero load growth for ExGen
• I am assuming zero rate base growth for the Utilities
• I am not modeling extra O&M synergies from the Constellation merger; management is claiming $500 million per year
• Exelon stands to benefit from greenhouse gas legislation since they have arguably the cleanest generating fleet in the country from a CO2 perspective. 

MANAGEMENT

They have relatively new senior leadership that was promoted internally.  Exelon’s main architect, John Rowe, retired a couple of years ago.  Chris Crane is the CEO.  He took over in 2011.  I admit he can sound a bit tongue-tied on their conference calls, but he is a hard core nuclear operator and that is exactly what I want.  That is a business where you cannot afford to screw up.  Chris has been with the company since 1998 and has been in the nuclear industry for over 25 years.

KEY RISKS

• Nuclear incident:  The risk that towers above others in terms of potential severity is a serious nuclear incident.  If something like Fukushima happens to Exelon you can basically assume Exelon’s common stock, holdco bonds, and ExGen bonds would only get back pennies on the dollar, if anything.
  
  The utility bonds should be okay because of creditor ringfencing, but they traded down on the knee-jerk reaction that might be an opportunity. 

• Cyberattack:  These are crucial infrastructure assets that could wreak havoc with the economy and environment if bad people managed to take control of them.  I don’t know anything about the degree to which their insurance coverage protects them against that risk.

• Recession:  the average business cycle is ~6 years between recessions and it’s been ~4 years since the last one.  Maybe we’re due.  If that happens between now and the 2016 time frame it could delay the realization of ExGen’s earnings upside.

Catalyst

• Nuclear incident
• Cyberattack
• Recession